1. Field of the Invention
The present invention relates to the generation of single or low order transverse mode coherent light from large volumes of asymmetric transverse cross section laser gain media to produce a filled-in, diffractively coupled output beam. This is accomplished by using a hybrid travelling wave unstable resonator. In general, it is well known that symmetric transverse cross section or asymmetric transverse cross section standing wave unstable resonators have superior transverse mode discrimination characteristics compared with stable standing wave resonators. Utilization of hybrid unstable ring resonators is shown in this disclosure to have features not obtainable using hybrid standing wave unstable optical cavities. For example, it is widely believed that hybrid negative branch standing wave unstable resonators uniquely achieve their high level of insensitivity to end-mirror misalignment because of the reversal of left-for-right that occurs at the focal plane of a concave mirror pair. Indeed, this is the reason that hybrid negative branch standing wave unstable cavities are said to be favored over hybrid positive branch standing wave cavities in some laser applications. However, this disclosure shows that the desirable intracavity left-for-right exchange feature and a low sensitivity to cavity misalignment can be achieved in a hybrid travelling wave cavity without having the deleterious characteristics associated with an intracavity focal plane. Likewise, because the cavity is travelling wave, spatial hole burning effects in some laser media are eliminated. Moreover, since the present invention is confined to a resonator unstable in one plane only, does not employ a bifurcated waveguide geometry, and is preferred to have all reflecting optics, the concept disclosed herein is scalable to very high average laser output powers. Accordingly, the general objects of the present invention are to provide novel and improved methods and apparatus of such character.
2. General Description of Unstable Resonator Prior Art
Since their first introduction [1] into the literature and their first systematic experimental and analytical investigation in 1965, unstable resonators have been applied to excimer, ionic, molecular, solid state, liquid state and free electronic laser media emitting over the spectral range from the ultraviolet to the infrared. In this initial paper, mode losses were found by an ad hoc geometric optical analysis to be independent of laser end mirror sizes and, while the cavity losses were found experimentally to be expectedly large, the thought was expressed that diffractive output coupling would be useful for transverse mode control. In 1967, a second paper [2] listed three general attributes of unstable resonators: “1) Unstable resonators can have large mode volumes even in very short resonators; 2) The unstable configuration is readily adapted to adjustable diffraction output coupling and 3) The analysis indicates that unstable resonators should have very substantial discrimination against higher-order transverse modes”. The first experimental evidence of high transverse mode discrimination in a laser with unstable resonator was reported in [3]. Over time, these three attributes have been confirmed experimentally and theoretically many times. In 1967 and 1969, the concept of standing wave unstable resonators was expanded to a confocal concept when several innovative unstable ring resonators were introduced for the first time [4] and briefly discussed [5,6] with the comment expecting “a new possibility of constructing unidirectional ring generators” [5, p 1002]. These publications [4-6] introduced unstable ring cavities for the first time with and without intracavity focal regions. These ring geometries were further explored [5,6] and for the first time it was found that “even though the losses of modes propagating in opposite directions are identical, their substantially different volumes in such cavities can obviously favor a unidirectional generation” [5, p 1002].
Although the CO2 laser medium was widely appreciated as one of a number of ideal candidates for the application of an unstable resonator system, it took nearly five years after the first introduction of unstable resonators [1] for details of such a resonator system to be reported in 1969 [7]. This work reported the use of a positive branch, confocal unstable resonator to generate a maximum cw output power of 22 watts and employed an annular coupler to generate a collimated fundamental output mode in the form of a near field annulus. Within a year of the first publication detailing the true diffraction losses in 1970 [8] of up to the first six lowest loss modes in a number of circular mirror, standing wave unstable resonator cavities, an unstable resonator system with a cw CO2 output power of 30 kW was reported [9].
In 1972, a patent for a unique confocal ring unstable resonator was filed [10]. Also, about this time a number of experimental CO2 laser studies were published exploring the very detailed characteristics of standing wave confocal unstable resonators [11], unidirectional symmetric confocal ring resonators [12], asymmetric confocal ring unstable resonators [13] and injection locking and regenerative amplification in standing wave and travelling wave unstable cavities [14]. Without exception, the experimental studies of the measured diffraction losses in confocal standing wave unstable cavities [11] was shown to be in complete agreement with the losses predicted by rigorous diffraction theory [8]. This agreement includes the details of the resonator loss characteristics near the transition between the lowest loss symmetric mode and the next lowest loss symmetric mode [Ref. 11, FIG. 17]. Likewise, unidirectional operation in travelling wave unstable resonators, first proposed mid-1968 [4] was achieved [12,13] as initially envisioned based only on the placement of the gain medium [5] within the unequal forward versus reverse mode volumes to favor one of the travelling wave directions. Moreover, the utility of travelling wave unstable resonators was shown to be a powerful resonator approach for applying to the concept of laser regenerative amplification [14]. In this case, unidirectional operation is shown to be strongly enhanced by the output mirror of the injection laser [Ref. 14, FIG. 35] which functions as the reversing mirror [Ref. 10, FIG. 24, element 24]. Unidirectional operation was shown to be easier to achieve in unstable optical ring regenerative amplifiers as compared to standing wave unstable resonators because no isolator is required [Ref. 10, FIG. 29].
All told, within a decade of first being introduced and analyzed the understanding of unstable optical resonators proceeded from an initial geometric optical approach [1] to a full iterative diffractive approach [8]. Along with this decade of theoretical work, CO2 output powers increased over the range from 20 W in an initial standing wave device [7] to eventually cw output levels speculated to be in the multi-hundred kW cw with an asymmetric ring unstable device design [12,13].
It is interesting to note historically that the original concept of unstable optical resonator [1] proposed by Siegman in 1965 was never submitted to any patent office for patenting. Perhaps this was due to a lack of a good diffractive analytical model for unstable resonators in the early days of discussion and development. Meanwhile sufficient practical utility of confocal unstable resonators was predicted in 1968 and demonstrated experimentally in 1969 independently in [7] and [15]. Due to these investigations the positive branch unstable confocal (telescopic) resonator innovation has been patented in Russia [16] with the priority date 18.03.1968, but for a long time remained unknown to the world laser community. The ring unstable resonator innovation [4], proposed in 1968 was not ever submitted to any patent office until 1972. Retrospectively, this may be due to general misunderstanding of how completely the reverse wave in the cavity can be suppressed. In 1972 unstable ring laser resonators were patented in [10] due to development of efficient concepts of unidirectional operation of lasers with such resonators. In any event, a contemporary review of unstable resonator works of that period can be found in [6,17] and a most thorough discussion of all these and other types of unstable resonators along with detailed references in [18,19].
Stable ring resonators were well known in the laser art of the late 1960's, having been introduced earlier for, among other things, applications requiring the sensing of physical rotation of objects in an inertial gravitational field [20]. For this application, the difference frequency between the forward and reverse ring waves was found to be proportional to the angular rotation rate of the ring laser system. Unstable ring resonators are distinguished from stable ring resonators in that the mode diameters in the forward and reverse directions are generally different in unstable ring cavities but the same in stable rings. This is the basis of one of the ways unidirectional operation [5-6] can be achieved through the use of an intracavity aperture. Also, suppression of one of the oscillation directions in either symmetric [12] or [6,12,13] asymmetric unstable ring resonators can be achieved by judicious placing of the intracavity gain medium. To accomplish this, one places the gain medium intracavity where the mode volume for one of the travelling waves is large and the other travelling waves is smaller [17, FIGS. 16,17]. In a near symmetric unstable ring cavity, the ratio of forward to reverse wave output power was measured to be nearly a factor of 20 [12, FIG. 6]. Another way of achieving unidirectional operation is through the use of a reversing mirror [10, FIG. 2] located outside the cavity. Indeed, the aspect of unidirectionality in both stable and unstable symmetric aperture resonators is central to the notion of achieving regenerative amplification without the introduction of an optical isolator between the master oscillator and regenerative amplifier [14, FIGS. 8, 29]. Likewise, in such applications as diverse plasma diagnostics [21] or analysis of laser spectral composition [22], ring geometries are highly advantageous and even essential. In all these applications, inventions or devices, universally and without exception, it should not be surprising to find that there is always some discussion of both directions of propagation in the ring geometry.
Obviously, in a travelling wave optical geometry, since the opposing directions of propagation exit the optical device in distinct and unique directions, to discuss only one of the propagating directions is equivalent to discussing only half of the optical problem. Indeed, without such a discussion it is impossible to even know with certainty which of the two counter propagating modes is being used for output or which direction the output will be extracted. Conversely, absent such a discussion of both propagation directions, such inventions or devices have to be considered to be fundamentally standing wave in nature and application.
Beyond the simple concept of directionality that one finds as the most distinguishing feature between stable ring resonators versus stable standing wave resonators, the differences between unstable ring resonators and stable ring resonators is far richer and more complex. For example, in a stable ring resonator, the mode diameters of the forward and reverse waves at any location in the resonator and the total mode volume of two waves is the same. In contrast, the mode diameters of the forward and reverse waves at any location in an unstable ring resonator and the total mode volumes of the two counter propagating waves are generally not the same.
For illustrative purposes, suppose an unstable ring resonator is both confocal and asymmetric. For this discussion, confocal refers to the fact that the design is such that either the forward or reverse wave is extracted from the resonator as a collimated output. Asymmetric in this case refers to the fact that the distance between the beam expansion optics is greater (or less than) the remaining portion of the perimeter. For such an asymmetric confocal case [10], the resonator is confocal in only one ring direction. Restated, “this kind of directional asymmetry can only be accomplished in a [unstable] ring resonator” [19, p 839 line 28,29]. Therefore to completely and unambiguously describe the modal properties of unstable ring resonators, they have to be discussed entirely separately from stable standing wave, stable traveling wave resonators and also standing wave unstable resonators.
Consequently, with respect to inventions claiming novelty by employing various types of symmetric aperture or hybrid unstable resonators, such inventions cannot be said to include unstable ring resonators unless the patent itself specifically includes a discussion as to how one of the unstable ring mode directions will be effectively suppressed. Likewise, some discussion should be presented as to what the shape of the unsuppressed travelling mode will be relative to the laser gain medium if it remains unsuppressed, since being unsuppressed, will represent a direction from which significant laser output power will be emitted. In this regard, U.S. Pat. No. 5,097,479 [23] conforms to this notion by describing the suppression of one of the travelling waves in a two mirror, bifurcated unstable ring resonator for application with a slab type CO2 laser medium. Likewise, U.S. Pat. No. 3,824,487 [10] conforms to this requirement since it discusses both the reverse wave and the accommodation of the unsuppressed wave to the large volume of gain medium. On the other hand, U.S. Pat. Nos. 4,719,639 [24] and 5,048,048 [25] fail in this regard and thus their utility is fundamentally self-limited to only hybrid standing wave unstable resonator geometries.
As herein disclosed, a laser with a travelling wave unstable resonator mode in one transverse dimension and either a waveguide or freespace gaussian mode in the orthogonal transverse dimension could be ideally suited for effectively coupling to any type of gain media with an elongated transverse cross section. This, of course, assumes that one of the unstable ring oscillation directions can be effectively suppressed. If so, this invention can be advantageously applied to excimer, ionic, molecular, solid state, liquid state or free electron laser media emitting over the spectral range from the ultraviolet to the infrared. Such media might be pumped by an RF, dc, e-beam, incoherent light, coherent light or free electron source, or any combination of these sources.
3. Description of RF Waveguide and Slab Laser Prior Art
While not limited thereto in its utility, the present invention is particularly well suited for applications in high power CO or CO2 lasers with rectangular discharge geometries. In general, a rectangular discharge geometry is one wherein the transverse discharge cross section is elongated and the discharge is established most typically in either the short transverse dimension (slab devices) or the long transverse discharge dimension (slice devices). A separate case exists for slice devices where the discharge can be established perpendicular to the elongated transverse aperture. In all these cases, the ratio of the long transverse dimension to the short transverse dimension is large and is such that the long dimension is able to support a travelling wave unstable resonator mode in this long transverse dimension. Because of the elongated transverse cross section, such lasers can advantageously employ optical resonators that have different functional and propagational characteristics in the two different transverse dimensions. For the first time an optical resonator of such a geometry was experimentally investigated in [26] with a slab-type Nd-glass laser. The cavity comprised one planar and one convex cylindrical mirror such that the resonator was unstable along the longer transverse dimension of the slab (240 mm) and equivalent to a Fabry-Perot resonator in the shorter dimension (20 mm). Subsequently similar resonators were termed hybrid [27]. Thus two types of hybrid resonators to which the present invention is particularly applicable is one where the field in the short transverse dimension is described by either i) a waveguide mode or ii) a freespace gaussian mode while that in the long dimension is functionally described by an unstable resonator mode.
In a preferred embodiment of this invention, the optical configuration disclosed could find utility in high-power collision cooled waveguide gas lasers as disclosed in the “slab” discharge geometry of '639 [24] and in the “slice” discharge geometry as disclosed in '663 [28].
A slab waveguide laser excited by transverse high-frequency electric discharge comprises a waveguide formed by the reflecting surfaces of two elongated electrodes disposed parallel and in opposition to one another. The electrodes are made of a material highly reflective to laser radiation, thus ensuring low radiation losses in the waveguide. The gap between the electrodes is filled by a gas gain medium, which is excited by a transverse electric discharge generated in the gas medium when high-frequency pump power is applied to the electrodes. Mirrors making up a standing wave laser resonator are disposed near both ends of the waveguide formed by the elongated electrode surfaces. Besides exciting the gas by the electric discharge and acting as the upper and lower walls for the optical waveguide, the electrodes play a role of cooling members and provide heat removal from the gain medium. To ensure adequate collisional heat removal from the discharge, the electrodes are made of a material with a high thermal conductivity. In addition, the gap between the electrodes is made small and does not usually exceed a few mm. The electric field in such a discharge is directed essentially perpendicular to the cooling member surfaces and is oriented essentially along the height of the gain medium cross-section. The typical examples of such lasers are CO2 [29,30], CO [31] and Xe [32] waveguide lasers with high-frequency excitation.
The “slice” laser discharge geometry disclosed in '663 [28] is also characterized by a gas gain medium excited by a high-frequency electric discharge and having an elongated transverse cross-section with a shorter and longer dimension. In distinction to the waveguide lasers the discharge region is defined by having the discharge electric field established perpendicular to the short transverse dimension. In slice lasers, the discharge is confined between a pair of closely spaced non-conductive cooling members. These cooling members are disposed in opposition to one another such that the gap between their surfaces opposed to each other is not only small enough to provide collision cooling of the gas filling the gap, but is suitable for guiding laser light between the non-conductive surfaces. Thus the discharge in the gas is excited by a system of electrodes disposed such that the electric field in the discharge chamber is essentially parallel to the surfaces of the cooling members, i.e. is directed transversely to the height (shorter dimension) of the gain medium cross-section. As disclosed in '663 [28] such “slice” lasers have a number of advantages in comparison with conventional waveguide lasers. These advantages include independent selection and optimization of discharge pressure and excitation frequency, the possibility of combined use of RF and dc discharge excitation sources and a moderated effect of boundary layers near the electrodes, among other virtues. Like slab devices, the slice discharge geometry is relevant to CO, CO2, Xe as well as other gas laser media.
Clearly, the small bore transverse RF discharge work with a microwave and RF geometry reported in 1980 [33] and the stripline geometry reported in 1984 [34] both predate the filing of the slab geometry '639 [24] in early 1987. On the basis of prior art, then, slab lasers were limited for patent purposes to only waveguide operation in the smaller of the two transverse discharge dimensions and only unstable resonator operation in the larger dimension. In the slab waveguide geometry the modes of light propagating along the opposing electrode surfaces are defined entirely by these surfaces and their mutual disposal. The slice geometry '663 [28] and '256 [35], on the other hand, was an entirely new transverse discharge arrangement when disclosed for the first time by patent application. Therefore, the slice geometry is one that a much wider set of hybrid resonator types can be applied. For example, with respect to the slice geometries [32,35] the term “light guide” as used in “slice” lasers has a broader sense than merely a waveguide mode. Thus light propagation in the slice geometry can be either waveguide as in a slab device or a case where the intracavity radiation propagates without any interaction with the slice chamber sidewalls. One such case is where the intracavity mode in the small transverse slice dimension does not touch the sidewalls and is best described as freespace gaussian mode in functionality. Such a case takes place for instance in stable resonators where the laser beam is confined by the resonator mirrors and does not touch the light guide sidewalls because the sidewalls are not a boundary condition for this type of intracavity modal propagation.
In the present disclosure the term “light guide” will be used in the more general sense taking in mind that it comprises all modes of light propagation from the waveguide mode to the free space propagation of gaussian beams.
To increase the volume of the gain medium in a conventional symmetric aperture waveguide laser and the laser output power, while at the same time maintaining a small electrode gap, a wide-aperture waveguide laser with plane-parallel elongated electrodes of width considerably in excess of the electrode gap was proposed in U.S. Pat. No. 4,719,639 [24]. The electrode surfaces reflecting laser radiation form in this laser the upper and lower walls of an optical waveguide of a large width, wherein the radiation propagating between the electrodes is confined by this waveguide only in the directions perpendicular to the electrode surfaces. The waveguide is open in the directions parallel to the electrode surfaces, and therefore the laser beam propagating along the waveguide can expand in these directions in both opposite senses as in free space. A convex and a concave confocal mirror making up a positive-branch standing wave unstable resonator with a magnification M>1 are disposed near the waveguide ends. In each transit from one mirror of this resonator to another and back, the laser beam expands by approximately M times in the two opposite directions wherein the beam is not confined by the electrode surfaces and can expand as in free space. To form only one output beam of a solid, i.e. filled-in cross section in a laser with such a resonator, the mirrors are usually disposed such that the axis of the unstable resonator formed by them is shifted to pass near one of the open sides of the electrode gap. The output laser beam is coupled out of the resonator from the other side of the electrode gap, more particularly, near the edge of the convex mirror overlapping only a part of the waveguide cross section. Such a “halved” configuration of the resonator allows formation at the laser exit, even for electrodes of a large width, of a beam of approximately rectangular solid cross section with a close to diffraction-limited divergence in each of the two transverse directions. The large electrode width provides excitation of a large gain medium volume and, as a result, a high output laser power.
It is known, however, that elongated discharge aperture lasers with a positive-branch unstable resonator having low magnification M are highly sensitive to resonator mirror misalignment, particularly to a change in their angular position in the plane parallel to the electrodes [25], as well as to wedge-type optical inhomogeneities in the same plane. Such inhomogeneities usually form in the laser gain medium under discharge pumping. This was not unexpected given earlier published works [6,7] with conventional unstable resonators. In slab or slice CO2 lasers magnification M does not, as a rule, exceed 1.2 to 1.5. Therefore, if special measures to increase the rigidity of the construction and to improve the gain medium optical homogeneity are not taken, mirror misalignment and gain medium inhomogeneities in such lasers result in a substantial deformation of the radiation mode structure. This will result in a sharp drop in output power, degradation of beam divergence, and an angular shift of the output beam, which cannot be tolerated in most applications. The need for taking measures to solve these problems will increase the laser cost.
It is well known to those of ordinary skill in the art of slab lasers that a shift of the positive-branch unstable resonator axis to one of the electrode gap sides will result in transition of the resonator configuration from a full resonator to the half configuration. However, it is not well appreciated that some intracavity radiation will still escape from the resonator on the side of the electrode gap where the shifted axis is situated. Thus, while such a shift in the resonator axis allows the formation of essentially one output beam of a solid cross section in place of two separate output beams, this advantageous feature comes at the cost of forcing some radiation from the side of the resonator where it cannot be combined into the useful output beam. This entails a loss of radiation to the cavity thus reducing overall laser efficiency. Worse, such radiation can be inadvertently coupled back into the desirable intracavity mode by stray reflections and force an undesirable higher order modes to compete for the full gain medium volume and thus the laser output.
To eliminate these difficulties, U.S. Pat. No. 5,048,048 [25] disclosed the use of a negative-branch linear unstable resonator with a magnification M<−1 in the wide dimension of the discharge aperture. The disclosed confocal geometry is formed by two concave mirrors with different radii of curvature having a common focal point inside the resonator. The confocal negative branch geometry can produce a one sided, filled in output beam as a result of the reversal of left for right which occurs at the confocal plane of the two concave resonator mirrors. In each pass through the focal point (beam focal waist), the laser beam propagating along the waveguide between the mirrors of the negative-branch unstable resonator becomes inverted in cross section, so that after passing through the focal point the rays of the beam that propagated on one side of the resonator axis (which is a common normal to the surfaces of both mirrors) will emerge on the other side of this axis. Due to the laser beam rays passing alternately on one and the other sides of the resonator axis, the misalignments caused by the resonator mirror angular shifts in the plane parallel to the electrodes become efficiently compensated for |M| on the order of 1.2 to 1.5, as are compensated efficiently also the wedge-type optical inhomogeneities in the gain medium, which makes the resonator only weakly sensitive to such misalignments and optical inhomogeneities [6,7].
To provide one-sided coupling of the radiation out of the negative-branch hybrid unstable resonator of '048[25], the size of one of the resonator mirrors is chosen such that the distances from the resonator axis to the opposite edges of this mirror in the plane parallel to the electrode reflecting surfaces differ by more than |M| times. The other resonator mirror is chosen large enough that it does not constrain beam expansion in the waveguide. On the next pass through the resonator, the radiation propagating along the resonator on the side of its axis opposite to the side on which the radiation is coupled out is reflected to the side where the beam exits, and is coupled out as the useful output beam. As a result, the laser output radiation is a solid cross-section beam which, despite the beam expanding in the resonator freely in two opposite directions, exits it on one of its sides only. Thus, the presence of a focal waist in a negative-branch unstable resonator reduces the passive losses of the radiation generated in the resonator compared to the positive-branch halved unstable resonator, in which beam expansion in two opposite directions brings about inevitably passive power losses from the resonator on its side opposite to the one where the useful output beam is coupled out.
However, because of the high local beam power density, the presence of a focal waist in the gain medium of a negative-branch unstable resonator may give rise to undesirable nonlinear effects in the gain medium and to gas breakdown, particularly in high-power pulsed lasers. Besides, the efficiency of use of the gain medium volume in such a resonator is lower than that in a positive-branch resonator because of the gain medium being nonuniformly filled by the beam. Reducing the laser dimensions, which is usually achieved by folding the optical axis of the resonator by means of an additional mirror, is also difficult in this arrangement, because the mirror placed into a negative-branch resonator to fold its axis will be too close to the focal waist to withstand the severe irradiation expected at high laser-power levels. Moreover, the mirrors of a negative-branch unstable resonator should have a large curvature; indeed, their curvature radii should be on the order of the distance between the mirrors. As a result, to reduce the effect of the curvature of these mirrors on field distribution over the waveguide height, i.e., along the normals to the electrode surfaces, one has to use mirrors of a complex shape, with different curvatures in the two mutually perpendicular directions, or to take special measures for wavefront matching, thus introducing additional losses in the resonator as disclosed in U.S. Pat. No. 5,123,028 [36]. Besides, if the electrode width is increased noticeably, the increase in the width of the large-curvature mirrors is accompanied by a fast growth of spherical aberrations entailing, in its turn, a substantial increase in the beam divergence, which also places an obstacle on the way to using negative-branch unstable resonators in high-power waveguide lasers. Indeed, while FIG. 4 of '048 [25] shows that there is an advantage to a negative branch configuration over that of a positive branch design, the power levels are noted to be relatively low. At the present time there is some indication that at the several kW output power level in CO2 slab lasers there is some significant beam steering that may be caused by gas heating or other non-linear effects at the common focal region of the confocal mirror pair.
U.S. Pat. No. 5,097,479 [23] proposed a wide-aperture waveguide gas laser with a positive-branch ring unstable resonator completed with means for forcing unidirectional oscillation in the resonator [23, FIG. 10, element 80]. This embodiment of a waveguide gas laser is pumped with high-frequency power applied between a pair of spaced electrodes. The ring resonator disclosed employs only two intracavity mirrors, that is why a complete ring resonator round trip must rely on a series of distributed reflections in the precisely curved bifurcated waveguide structure. In this split bi-waveguide structure an unstable ring optical resonator is formed with a closed axial contour to permit the extraction of an output beam with a solid cross section.
The ring resonator depicted in '479 [23] is formed by optically combining two precisely curved branches of two adjacent optical waveguide structures into a single optical unit. The precisely curved, bifurcated waveguide structure halves are coupled optically together by means of a pair of mirrors disposed at the ends so as to direct the laser beam impinging on the mirror from one waveguide branch into the other waveguide branch. As a result, each of the mirrors turns the beam striking it in the plane transverse to the electrode surfaces. Thus, a pair of mirrors and two curved waveguide branches form in this laser a ring resonator with a closed axial contour lying in the plane which crosses the electrode surfaces essentially at right angles and faces with its opposite sides the open side ends of the waveguide. The ring resonator formed in this way in this plane provides a compact laser design, because the height of these waveguide branches is small compared to their width and length. However, since it is impossible to form a travelling wave resonator of any kind using only two mirrors, the ring resonator optical circuit must rely on a proper, precise and equal curvature of both waveguide branches. Clearly, the continuous reflections in the two curved bifurcated waveguide branches will add considerable intracavity optical loss and significant mechanical complexity to the laser fabrication process as well. This is especially evident when it is remembered that one of the electrodes in a RF pumped slab laser must be at an elevated RF potential. On balance, the apparent simplicity of using only two intracavity ring mirrors must be weighed against the complexity arising from the requirement of achieving a very low distributed optical reflection loss along both curved waveguide branches. This low distributed loss must be achieved at the same time as keeping the elevated potential RF electrode from shorting out to the grounded RF electrode. Moreover, to provide one-sided beam extraction from said positive-branch ring unstable resonator, the axial contour of this resonator is shifted such that it passes near one of the open sides of the electrode gap. As can be seen from the optical diagram of '479 [23], this hybrid ring resonator approach cannot be applied to a single plane waveguide or a single plane guided wave structure since it only uses two intracavity optical elements. Furthermore, because of the bifurcated waveguide structure, the ring resonator cannot be made asymmetric.
Because the traveling wave unstable resonator of '479[23] has no focal waists inside the cavity, one-sided diffractive output extraction from the laser is provided only by shifting the axial contour to one of the open sides of the electrode gap. Consequently, this laser suffers from all the disadvantages discussed in relation to slab lasers with a positive branch, halved linear unstable system. Among these disadvantages are a high sensitivity to resonator mirror misalignment and wedge-type inhomogeneities in the gain medium, as well as passive radiation losses on the resonator side to which the resonator axial contour is shifted occurring when extracting the useful radiation in the form of one beam. If anything, the bifurcated waveguide, traveling wave unstable resonator system of '479[23] adds, rather than eliminates, complexities of the other single slab hybrid resonator slab devices.